EDP Sciences logo
Subscriber Authentication Point
Search
Menu
Home All issues Volume 699 (July 2025) A&A, 699 (2025) L1 Full HTML
Open Access
Issue
A&A
Volume 699, July 2025
Article Number L1
Number of page(s) 9
Section Letters to the Editor
DOI https://doi.org/10.1051/0004-6361/202555495
Published online 25 June 2025

© The Authors 2025

Licence Creative CommonsOpen Access article, published by EDP Sciences, under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

This article is published in open access under the Subscribe to Open model. Subscribe to A&A to support open access publication.

1. Introduction

Classical B emission-line stars (Be stars, hereafter) are rapidly rotating main-sequence (MS) B-type stars surrounded by ionised, gaseous discs. The exact mechanism behind the formation of these discs, commonly known as the ‘Be phenomenon’, remains elusive. Various mechanisms have been proposed to explain how the material is expelled from the Be star into the disc, including rotation, non-radial pulsation, magnetic fields, and the presence of binary companions (Rivinius et al. 2013; Bodensteiner et al. 2020; Mathew et al. 2008).

The binary formation pathway suggests the crucial role of a companion star that fills its Roche lobe, leading to the formation of a circumstellar disc around the Be star. Based on simulations of binary star populations, de Mink et al. (2013) suggested that rapidly rotating massive stars could be spun up either through mass transfer or mergers. Additionally, de Mink et al. (2014) concluded that approximately 30% of all MS B-type stars are products of dynamical interactions occurring among the stars in clusters. Binary population synthesis studies specific to Be stars (Pols et al. 1991; van Bever & Vanbeveren 1997; Shao & Li 2014, 2021) also suggest that at least a significant fraction, if not all, are likely the products of binary interactions.

A series of studies conducted by Peters et al. (2008, 2013, 2016) also support the existence of invisible companions that play a role in the formation of discs in Be stars. Hot companions of these stars are faint in optical and infra-red (IR) but are easily detectable in the far-ultraviolet (FUV) as they emit most of their flux at short wavelengths. Recently, Wang et al. (2021) increased the sample of known Be+sdO binaries.

The majority of the Be-companions are hot subluminous stars of subdwarf O, B-type (sdOB) stars (Klement et al. 2022). These companions can subsequently evolve into white dwarfs (WDs) or neutron stars (NSs), potentially leading to the formation of Be/X-ray binaries (BeXRBs). Some of these post-mass transfer (post-MT) systems have already been identified as Be-NS or Be-WD systems through X-ray outbursts. To detect stripped stars next to an optically bright companion, Götberg et al. (2018) recommended systematic searches for their UV excess. This clearly showcases the importance and need to employ UV imaging to shed more light on their binary formation pathway. So far, no Be+sdOB system has been detected in star clusters.

Young open clusters (OCs) with an age of less than a hundred million years are ideal laboratories for the study of Be stars among the B-type stars and the role of binarity in shaping their properties. In the study by Mathew et al. (2008) covering emission line stars in 42 young OCs, NGC 663 stood out as one of the clusters with a large number of Be stars. NGC 663 is a young OC located in the constellation Cassiopeia. Pigulski et al. (2001a) estimated an age of 20−25 Myr and a distance of 2.1 kpc. Recently, Dias et al. (2021) estimated the cluster distance to be 2.353 kpc, log(t) = 7.403 yr, [Fe/H] = −0.125 dex, and AV = 2.268 mag, which are adopted in this work.

This study investigates the ultraviolet properties of Be stars in NGC 663 to check for the presence of hot companions and reports the first detection of binary Be stars with subdwarf companions in OCs, providing evidence for such a formation pathway for Be phenomena in OCs.

2. Data reduction and analysis

We used archival data of UVIT onboard AstroSat (L1 data from the AstroSat data archive1) and processed them using CCDLAB Postma & Leahy (2017). A detailed description of UVIT and its calibration can be found in Tandon et al. (2017, 2020). NGC 663 was observed with UVIT on 12 August, 2017, (Proposal Id: A03-079) in three FUV (F148W, F154W, F169M) and two NUV filters (N263M and N279N). The CCDLAB corrects for satellite drift, flat field, distortion, fixed pattern noise, and cosmic rays. The orbit-wise images are then co-aligned and merged to get the final science-ready images. Details of the UVIT observations of NGC 663 are tabulated in Table A.1. Figure A.1a shows a composite two-colour image of NGC 663 (F148W and N279N).

3. Photometry

Aperture and point spread function (PSF) photometry were carried out using the DAOPHOT package in IRAF/NOAO Stetson (1987). The model PSF was constructed using bright, isolated stars and then applied to all the detected stars using the ALLSTAR task to obtain PSF-fitted magnitudes. The PSF, aperture, and saturation corrections were subsequently applied to obtain the instrumental magnitudes. We note that many bright stars in the cluster were too saturated to derive reliable magnitudes.

The details of the saturation correction are described in Tandon et al. (2017). Using zero-point (ZP) magnitudes reported in the calibration paper (Tandon et al. 2020), the derived instrumental magnitudes were calibrated into the AB magnitude system. By limiting the PSF fitted error to ∼0.2 mag (see Fig. A.1b), stars detected down to ∼21 mag in the F148W, F154W, N236M filters, and down to ∼19 mag in the F169M and N279N filters, were selected for further analysis. We used the reddening value of E(B − V) = 0.7 mag (Dias et al. 2021) and RV = 3.1 from Whitford (1958) for the Galaxy. Given the differential extinction within the cluster reported by Yadav & Sagar (2001), we used a range in the AV values with the corresponding extinction in the respective filters tabulated in Table A.1 to overlay the isochrones.

4. Be stars in NGC 663

Sanduleak (1979) found 27 Hα emission stars, mostly earlier than the B5 spectral type. This number was later updated to 24, with 12 stars showing variability in Hα emission (Sanduleak 1990). Subsequently, Pigulski et al. (2001b) identified 26 Be stars, the majority of which are between B0-B3. Later, using the slitless spectroscopy method, Mathew et al. (2008) detected several new Be stars, increasing the total count in the cluster to 31, within the B0-B8 range. Among the 31 known Be stars, Yu et al. (2015) re-identified 15 stars and discovered four new Be stars. They concluded that NGC 663 contains a total of 34 Be stars. Of these, 23 stars were detected by UVIT which are analysed in this study.

5. Spectral energy distributions of Be stars

To derive the stellar parameters of the Be stars detected with UVIT, we constructed their spectral energy distributions (SEDs) using the Virtual Observatory SED Analyser (VOSA; Bayo et al. 2008). Following the construction of the SEDs, best-fit parameters such as effective temperature (Teff), luminosity (L), and radius (R) are derived by performing a reduced chi-square (χred2) minimisation fitting, which compares the observed fluxes to the synthetic photometric fluxes. The SED fitting technique is described in more detail in Rani et al. (2021). In addition to the χred2 value, VOSA computes two additional parameters, Vgf and Vgfb, which correspond to modified χred2 values and are used to evaluate the goodness of the fit with minimum fractional residual, especially when the observational flux errors are minimal. For an SED fit to be reliable, the Vgfb value must be less than 15 (Rebassa-Mansergas et al. 2021). To estimate errors in the derived parameters, VOSA makes use of Markov chain Monte Carlo (MCMC) approach.

We used Kurucz stellar atmospheric models (Castelli & Kurucz 2003) to compute synthetic SEDs across a wavelength range extending from FUV to the IR, encompassing observed photometric data points. These models provide the grid of parameters such as Teff, metallicity ([Fe/H]), and surface gravity log g. We fixed metallicity, [Fe/H] of −0.5 dex, close to the cluster metallicity (Paunzen et al. 2010). We adopted a range of log g values from 3−5 dex, as Be stars span a range of evolutionary phases (Peters 1976) and kept Teff as a free parameter (3500−50 000 K) in the case of the Kurucz models. We combined three FUV and two NUV UVIT photometric data points with Gaia EDR3 (3 passbands), SLOAN/SDSS (5 passband), 2MASS (3 passbands), and WISE (4 passbands) to construct the observed SEDs. First, we fitted the observed SED, incorporating all the data points, with a single model SED in order to check the overall fit. Out of 23 stars, four stars are well fitted with a single model SED fit, and 19 stars show a significant deviation (> 50%) from the model at shorter wavelengths (FUV region). To further confirm the UV excess, we repeated the single model SED fit by excluding the data points with wavelengths less than 3000 Å. After obtaining the stellar parameters, we employed the VOSA binary fit tool to fit the hotter part of the SED for 19 stars. The VOSA binary fitting tool simultaneously fits the hotter and cooler components of a binary system by assuming that the observed flux is a linear combination of two different models. The Kurucz and TMAP atmospheric models (Rauch & Deetjen 2003) are used to fit the hotter component. The TMAP model grid spans a range of atmospheric parameters, with Teff between 50 000 and 190 000 K, surface gravities (log g) between 5 and 9 dex, and a hydrogen mass fraction of zero. For the star GG 101, we were unable to obtain a satisfactory fit using either the Kurucz or TMAP models. Therefore, we adopted the Levenhagen model (Levenhagen et al. 2017), which covers Teff values from 17 000 to 100 000 K and log g between 7 and 9.5 dex.

Out of the 19 stars with UV excess, 16 stars are well fitted with a double-component SED, except for three stars: D01 034, GG 109, and SAN 28. We show the best single fit and binary fit SED of Be star GG95 (upper panel), and VES 619 (lower panel) in Fig. 1. The best-fit parameters of the 23 Be stars (seven single and 16 binary) derived from the best SED fits with the lowest χred2 and Vgfb values are listed in Table B.1. We obtained Vgfb values below 15 for all of the Be stars, indicating good SED fits and confirming the reliability of all the derived fundamental parameters. The remaining best-fit single and double-component SEDs are provided in Appendix B. Since classical Be stars exhibit excess emission in the near-UV (NUV) and IR due to free-bound and free-free processes occurring in circumstellar discs, respectively (Goraya 1986), these mechanisms are not efficient at producing significant flux in the FUV regime. Also, note that chromospheric activity cannot account for this UV excess, as B-type stars do not exhibit such activity due to the absence of convective envelopes (Hall 2008). Thus, the observed FUV excesses in our sample suggest an additional hot companion. Furthermore, to assess the evolutionary stage and nature of hot companions, we place them in the Hertzsprung-Russell diagram (HRD), as discussed in the subsequent section.

thumbnail Fig. 1.

Upper panel: Single-fit SED of Be star GG 95, with the Kurucz model (green line) fitted to observed fluxes (red). Lower panel: SED of Be star VES 619, fitted with a combined Kurucz (dash-dotted blue line, cool) and TMAP (dash-dotted brown line, hot) model. The composite fit here is shown with a solid green line. Best-fit parameters and model details are shown above each plot.

6. Discussion

This is the first study to trace possible binary companions of Be stars in young clusters. NGC 663 is known to harbour a large number of Be stars for a very long time. These stars also have a large range in spectral type, leading to the inference that the ‘Be-phenomenon’ is operational across the B spectral type in this cluster. Modelling the UV excess and the multi-wavelength SED of 23 Be stars has led to the detection and characterisation of binary companions in 16. The hot companions, as shown in HRD, (Fig. 2) belong to the sdOB spectral types and appear to be similar to those found in the field. This study, therefore, provides the first evidence of a binary pathway to Be formation in a young Galactic cluster. NGC 663 has at least 69.5% of Be stars formed through binary mass-transfer, suggesting that a majority of Be stars in NGC 663 are formed via binary mass transfer.

thumbnail Fig. 2.

HR diagram illustrating UVIT-identified Be stars and their hot companions. The cluster isochrone (dashed black line), WD evolutionary tracks (dash-dotted black line, 0.3−0.5 M⊙, from right to left), and MIST tracks (solid black line, 3.0−8.2 M⊙, bottom to top) are shown for comparison. Cool and hot components of binary Be stars are marked in green and brown, and single-fit Be stars are in blue. Field Be+sdO systems from Wang et al. (2021) are displayed in purple, and model stripped stars with masses of 0.58, 0.66, 0.74, 0.85, 0.97, 1.11, 1.27, 1.43, 1.63, 1.88, 2.17, 2.49, and 2.87 M⊙ from Götberg et al. (2018) are shown as golden circles (bottom to top).

Wang et al. (2021) detected spectral signatures of the sdO companions in ten out of 13 Be stars, and a comparison of their temperatures and radii with evolutionary tracks indicates that the sdO stars occupy the relatively long-lived, He-core burning stage. The companions identified to Be-stars in NGC 663 also occupy a very similar space in the HR diagram (see Fig. 17 of Wang et al. 2021), suggesting that they are sdO stars in the long-lived He-core burning phase.

The cluster NGC 663 is 25 Myr old and therefore has a turn-off mass of ∼9.2 M⊙. Recently, Cavallo et al. (2024) estimated the cluster’s age to be 39 Myr, which corresponds to a turn-off mass between ∼7−8 M⊙. All 16 sdOBs likely formed within 25 Myr from a primary star mass of 7−9 M⊙ in a binary system. Even though the initial mass is similar, the end mass appears to be different. When we compare the L and Teff of the detected sdOBs with the models from Götberg et al. (2018), we find that the mass of the end products likely ranges from 0.6−2.8 M⊙. The primary stars would have lost different amounts of mass during mass transfer, possibly due to different orbital parameters. The Be stars, which are the cooler companion of the sdOBs, also have a mass range of approximately 2−8 M⊙, suggesting that the Be binaries have a wide range of primary-to-secondary star mass-ratios. This likely explains the appearance of Be phenomenon across the B spectral type in this cluster. The dynamical environment of the cluster also likely plays a role in binary evolution, as this can alter binary orbits. The orbital information for these systems is currently unknown, except that many are known to be photometric and/or spectroscopic variables. The binary Be systems in this cluster are therefore very important for investigating binary mass transfer mechanisms in massive stars. The sdOB companions detected here put strong constraints on the mass transfer mechanisms that are crucial to theoretical models such as those presented in Götberg et al. (2018).

The high-resolution spectroscopic studies of the identified Be binary systems in NGC 663 need to be performed to confirm the evolutionary status as well as to explore the atmospheric abundances of the sdOBs. Photometric and spectroscopic variability of these systems also need to be studied systematically to identify orbital parameters. In the future, we plan to extend this study to more young clusters harbouring Be stars.

7. Summary

The main results from this work can be summarised as follows:

  • We utilised UVIT/AstroSat observations in 3 FUV and 2 NUV filters to characterise 23 Be stars in the young OC NGC 663.

  • We estimated atmospheric parameters such as Teff, Ls, and Rs of the Be stars by fitting multi-wavelength SEDs with a single model spectrum. Among the 23 Be stars considered, 19 exhibited significant UV excess.

  • Sixteen Be stars were well fitted with a double-component SED, indicating the presence of a hot companion. The nature of these hot companions was assessed by placing them on the HR diagram and comparing their position with the theoretical evolutionary tracks or previously observed field stars of similar class. The positions of the detected hot companions coincide with the location of stripped stars (Wang et al. 2021), suggesting that the companions are likely stripped stars belonging to the sdOB type in the long-lived He-core burning phase.

  • The estimated masses of the hot companions range from 0.6 to 2.8 M⊙, based on the models of Götberg et al. (2018), while the Be stars have masses between 2.0 and 8.0 M⊙, derived from comparison with MIST evolutionary tracks, with many of them being slightly evolved from the MS.

  • This study establishes the binary pathway for the formation of Be stars in young clusters, with 69.5% of Be stars in NGC 663 found in binary systems. Be binary systems are interesting targets for further high-resolution spectroscopic studies to probe their evolutionary status and to gain insights into the abundances of the sdOB stars.

Acknowledgments

We thank the referee for providing the productive report that improved the quality of our manuscript. AS acknowledges support from the SERB Power Fellowship. SN and SR acknowledges support from StarDance: the non-canonical evolution of stars in clusters (co-funded by the European Union, ERC-2022-AdG, Grant Agreement 101093572, PI: E. Pancino). SN also extends gratitude to Blesson Mathew, Sipra Hota, Ashish Devraj, Akhil Krishna and Vikrant Jadhav for the fruitful discussions. This publication utilises data from the AstroSat mission’s UVIT, archived at the Indian Space Science Data Centre (ISSDC). The UVIT project is a collaboration between IIA Bengaluru, IUCAA Pune, TIFR Mumbai, several ISRO centres, and CSA. This research also made use of VOSA, developed under the Spanish Virtual Observatory project, supported by the Spanish MINECO through grant AyA2017-84089. Software. Topcat (Taylor 2011), Matplotlib (Hunter 2007), NumPy (van der Walt et al. 2011), Scipy (Oliphant 2007; Millman & Aivazis 2011), Astropy (Pigulski et al. 2001c; Astropy Collaboration 2018), and Pandas (McKinney et al. 2010).

References

  1. Angelo, M. S., Santos, J. F. C., Maia, F. F. S., & Corradi, W. J. B. 2022, MNRAS, 510, 5695 [CrossRef] [Google Scholar]
  2. Astropy Collaboration (Price-Whelan, A. M., et al.) 2018, AJ, 156, 123 [Google Scholar]
  3. Bayo, A., Rodrigo, C., Barrado Y Navascués, D., et al. 2008, A&A, 492, 277 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  4. Bodensteiner, J., Shenar, T., & Sana, H. 2020, A&A, 641, A42 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  5. Castelli, F., & Kurucz, R. L. 2003, IAU Symp., 210, A20 [Google Scholar]
  6. Cavallo, L., Spina, L., Carraro, G., et al. 2024, AJ, 167, 12 [NASA ADS] [CrossRef] [Google Scholar]
  7. Choi, J., Dotter, A., Conroy, C., et al. 2016, ApJ, 823, 102 [Google Scholar]
  8. Cordoni, G., Casagrande, L., Yu, J., et al. 2024, MNRAS, 532, 1547 [CrossRef] [Google Scholar]
  9. de Mink, S. E., Langer, N., Izzard, R. G., Sana, H., & de Koter, A. 2013, ApJ, 764, 166 [Google Scholar]
  10. de Mink, S. E., Sana, H., Langer, N., Izzard, R. G., & Schneider, F. R. N. 2014, ApJ, 782, 7 [Google Scholar]
  11. Dias, W. S., Monteiro, H., Moitinho, A., et al. 2021, MNRAS, 504, 356 [NASA ADS] [CrossRef] [Google Scholar]
  12. Gaia Collaboration 2022, VizieR Online Data Catalog: I/355 [Google Scholar]
  13. Goraya, P. S. 1986, MNRAS, 222, 121 [Google Scholar]
  14. Götberg, Y., de Mink, S. E., Groh, J. H., et al. 2018, A&A, 615, A78 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  15. Granada, A., Jones, C. E., Sigut, T. A. A., et al. 2018, AJ, 155, 50 [NASA ADS] [CrossRef] [Google Scholar]
  16. Hall, J. C. 2008, Liv. Rev. Sol. Phys., 5, 2 [Google Scholar]
  17. Hunt, E. L., & Reffert, S. 2023, VizieR Online Data Catalog: J/A+A/673/A114 [Google Scholar]
  18. Hunter, J. D. 2007, Comput. Sci. Eng., 9, 90 [NASA ADS] [CrossRef] [Google Scholar]
  19. Klement, R., Baade, D., Rivinius, T., et al. 2022, ApJ, 940, 86 [NASA ADS] [CrossRef] [Google Scholar]
  20. Levenhagen, R. S., Diaz, M. P., Coelho, P. R. T., & Hubeny, I. 2017, ApJS, 231, 1 [NASA ADS] [CrossRef] [Google Scholar]
  21. Mathew, B., Subramaniam, A., & Bhatt, B. C. 2008, MNRAS, 388, 1879 [NASA ADS] [CrossRef] [Google Scholar]
  22. McKinney, W. 2010, in Proceedings of the 9th Python in Science Conference, eds. S. van der Walt, & J. Millman, 56 [Google Scholar]
  23. Millman, K. J., & Aivazis, M. 2011, Comput. Sci. Eng., 13, 9 [NASA ADS] [CrossRef] [Google Scholar]
  24. Oliphant, T. E. 2007, Comput. Sci. Eng., 9, 10 [NASA ADS] [CrossRef] [Google Scholar]
  25. Paunzen, E., Heiter, U., Netopil, M., & Soubiran, C. 2010, A&A, 517, A32 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  26. Peters, G. J. 1976, IAU Symp., 70, 69 [Google Scholar]
  27. Peters, G. J., Gies, D. R., Grundstrom, E. D., & McSwain, M. V. 2008, ApJ, 686, 1280 [Google Scholar]
  28. Peters, G. J., Pewett, T. D., Gies, D. R., Touhami, Y. N., & Grundstrom, E. D. 2013, ApJ, 765, 2 [Google Scholar]
  29. Peters, G. J., Wang, L., Gies, D. R., & Grundstrom, E. D. 2016, ApJ, 828, 47 [Google Scholar]
  30. Pigulski, A., Kopacki, G., & KoÅ‚aczkowski, Z. 2001a, A&A, 376, 144 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  31. Pigulski, A., Kopacki, G., & Kolaczkowski, Z. 2001b, VizieR Online Data Catalog: J/A+A/376/144 [Google Scholar]
  32. Pigulski, A., Kopacki, G., & Kolaczkowski, Z. 2001c, A&A, 376, 144 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  33. Pols, O. R., Cote, J., Waters, L. B. F. M., & Heise, J. 1991, A&A, 241, 419 [NASA ADS] [Google Scholar]
  34. Postma, J. E., & Leahy, D. 2017, PASP, 129, 115002 [Google Scholar]
  35. Rani, S., Pandey, G., Subramaniam, A., et al. 2021, ApJ, 923, 162 [NASA ADS] [CrossRef] [Google Scholar]
  36. Rauch, T., & Deetjen, J. L. 2003, ASP Conf. Ser., 288, 103 [NASA ADS] [Google Scholar]
  37. Rebassa-Mansergas, A., Solano, E., Jiménez-Esteban, F. M., et al. 2021, MNRAS, 506, 5201 [NASA ADS] [CrossRef] [Google Scholar]
  38. Rivinius, T., Carciofi, A. C., & Martayan, C. 2013, A&ARv, 21, 69 [Google Scholar]
  39. Sanduleak, N. 1979, AJ, 84, 1319 [Google Scholar]
  40. Sanduleak, N. 1990, AJ, 100, 1239 [Google Scholar]
  41. Shao, Y., & Li, X.-D. 2014, ApJ, 796, 37 [Google Scholar]
  42. Shao, Y., & Li, X.-D. 2021, ApJ, 908, 67 [NASA ADS] [CrossRef] [Google Scholar]
  43. Stetson, P. B. 1987, PASP, 99, 191 [Google Scholar]
  44. Tandon, S. N., Subramaniam, A., Girish, V., et al. 2017, AJ, 154, 128 [NASA ADS] [CrossRef] [Google Scholar]
  45. Tandon, S. N., Postma, J., Joseph, P., et al. 2020, AJ, 159, 158 [Google Scholar]
  46. Taylor, M. 2011, Astrophysics Source Code Library [record ascl:1101.010] [Google Scholar]
  47. van Bever, J., & Vanbeveren, D. 1997, A&A, 322, 116 [NASA ADS] [Google Scholar]
  48. van der Walt, S., Colbert, S. C., & Varoquaux, G. 2011, Comput. Sci. Eng., 13, 22 [Google Scholar]
  49. Wang, L., Gies, D. R., Peters, G. J., et al. 2021, AJ, 161, 248 [Google Scholar]
  50. Whitford, A. E. 1958, AJ, 63, 201 [NASA ADS] [CrossRef] [Google Scholar]
  51. Yadav, R. K. S., & Sagar, R. 2001, MNRAS, 328, 370 [Google Scholar]
  52. Yu, P. C., Lin, C. C., Chen, W. P., et al. 2015, AJ, 149, 43 [Google Scholar]

Appendix A: Colour-magnitude diagrams

Hunt & Reffert (2023) present the Proper motion (PM) membership study of 7167 star clusters using Gaia DR3 catalogue (Gaia Collaboration 2022), including NGC 663, and they found 1081 stars as PM members of the cluster with membership probability (PPM) more than 50%. To identify optical counterparts of UVIT detections, we cross-matched the Hunt & Reffert (2023) catalogue (PPM > 50%) with sources detected in both FUV and NUV images. The PPM of the 23 Be stars are tabulated in Table B.1.

Figure A.2 shows the optical, FUV-optical, and NUV-optical colour-magnitude diagrams (CMDs) along with overlaid MIST isochrones taken from Choi et al. (2016) generated for both UVIT and Gaia DR3 filters for the cluster parameters (as mentioned in the caption). We used isochrones with a metallicity of [Fe/H] = -0.125 dex and Z = 0.003110, excluding initial rotation effects. The optical magnitudes were converted from Vega to AB magnitude system using appropriate conversion factors provided in Gaia documentation2.

We cross-matched UVIT-detected stars with Be stars catalogued in Yu et al. (2015), identifying 23 stars. Of these, 22 are detected in F148W, F154W, and F169M, 1 in N263M, and 23 in N279N, all with membership probabilities exceeding 50%. All identified Be stars are represented by green triangles, cyan rectangles and blue pentagons in the optical CMD, as illustrated in Fig. A.2. A significant spread is observed along the MS, especially at its bright end spanning ∼0.45 mag in BP−RP colour. The FUV-optical and NUV-optical CMDs also show the spread at the bright end of the MS, but FUV and NUV optical colours span ∼2 mag larger than optical colour mainly due to the sensitivity of the UV-optical colour to Teff. Also, the turn-off region of NGC 663 appears slightly broadened, partially influenced by the presence of Be stars. Extended main-sequence turn-offs (eMSTOs), likely caused by enhanced stellar rotation, have been observed in OCs with ages between approximately 50 Myr and 2 Gyr (see (Cordoni et al. 2024) and references therein). However, NGC 663 is younger than the typical eMSTO clusters, and the spread at the turn-off appears to be prominent; therefore, this cluster may be the youngest to show the eMSTO phenomenon. As this cluster is known to have a differential reddening, we also adopt an extinction range from Av = 2.17 to Av = 2.9 in all CMDs.

Since the Be stars have a high rotational velocity, their Teff are affected leading to fast rotators on the red edge of the MS, whereas slow rotators in its blue side. To assess the impact of Teff on UV-optical colour (which is more sensitive to Teff), colour-colour diagrams (CCDs) are constructed using the FUV and NUV filters. The linear plot of (Gbp-Grp) against (F154W-G), (F169M-G), and (N279-G) are shown in Fig. A.2. In all CCDs, optical colour (Gbp-Grp) spans a 0.6 mag range, whereas the FUV-G colour spans about 3 mag, a similar trend can be seen in the case of the NUV-G (∼2.5 mag) colour diagram. We note that some Be stars show bluer UV-optical colour, suggestive of a possible UV excess, as they are bluer in the UV-optical colour.

Table A.1.

UVIT observations and different extinction values adopted for NGC 663.

thumbnail Fig. A.1.

Left: UVIT two-colour image of NGC 663. Right: PSF-fit magnitude errors.

thumbnail Fig. A.2.

Top panel: Optical CMD of NGC 663, based on Gaia DR3 photometry (first panel). The red points represent the optical counterparts (Gaia DR3) of the UVIT FUV/NUV detections. Be stars with UV excess, Be stars without UV excess, and Be stars showing UV excess but not fitted with a double component SED are shown as green triangles, cyan squares, and blue pentagons, respectively. The MIST isochrones corresponding to an age of 25 Myr and metallicity [Fe/H] = -0.125 dex are overplotted as solid black lines for various values of AV, ranging from 2.17 to 2.9 (from left to right), assuming a reddening of E(B-V) = 0.7 mag and a distance modulus of (m-M)V = 11.86 mag. The second and third panels in the upper panel show the FUV-optical and NUV-optical CMDs created using the F154W and N279N passbands, respectively, for the confirmed members of the cluster. Median error bars are displayed in grey on the left side of FUV and NUV-optical CMDs. Bottom panel: CCDs in the FUV-F154W vs F154W, FUV-F169M vs F169M, and NUV-N279N vs N279N planes, respectively. The meaning of the symbols shown is provided in the top-left panel of the first column.

Appendix B: SED fits of Be stars

The details of the SED fitting are elaborated in Sect. 5. The SED fits, single as well as binary, of the rest of the Be stars are shown in Fig. B.1

Table B.1.

Fitting parameters for 23 detected Be stars in NGC 663.

thumbnail Fig. B.1.

Best SED fits, including single and binary model fits, of the remaining Be stars. The SEDs illustrated in the uppermost panel consist of three stars showing no UV excess. They are well fitted with a single model SED. The symbols and models are the same as in Fig. 1.

thumbnail Fig. B.1.

Continued.

All Tables

Table A.1.

UVIT observations and different extinction values adopted for NGC 663.

In the text
Table B.1.

Fitting parameters for 23 detected Be stars in NGC 663.

In the text

All Figures

thumbnail Fig. 1.

Upper panel: Single-fit SED of Be star GG 95, with the Kurucz model (green line) fitted to observed fluxes (red). Lower panel: SED of Be star VES 619, fitted with a combined Kurucz (dash-dotted blue line, cool) and TMAP (dash-dotted brown line, hot) model. The composite fit here is shown with a solid green line. Best-fit parameters and model details are shown above each plot.
In the text
thumbnail Fig. 2.

HR diagram illustrating UVIT-identified Be stars and their hot companions. The cluster isochrone (dashed black line), WD evolutionary tracks (dash-dotted black line, 0.3−0.5 M⊙, from right to left), and MIST tracks (solid black line, 3.0−8.2 M⊙, bottom to top) are shown for comparison. Cool and hot components of binary Be stars are marked in green and brown, and single-fit Be stars are in blue. Field Be+sdO systems from Wang et al. (2021) are displayed in purple, and model stripped stars with masses of 0.58, 0.66, 0.74, 0.85, 0.97, 1.11, 1.27, 1.43, 1.63, 1.88, 2.17, 2.49, and 2.87 M⊙ from Götberg et al. (2018) are shown as golden circles (bottom to top).
In the text
thumbnail Fig. A.1.

Left: UVIT two-colour image of NGC 663. Right: PSF-fit magnitude errors.
In the text
thumbnail Fig. A.2.

Top panel: Optical CMD of NGC 663, based on Gaia DR3 photometry (first panel). The red points represent the optical counterparts (Gaia DR3) of the UVIT FUV/NUV detections. Be stars with UV excess, Be stars without UV excess, and Be stars showing UV excess but not fitted with a double component SED are shown as green triangles, cyan squares, and blue pentagons, respectively. The MIST isochrones corresponding to an age of 25 Myr and metallicity [Fe/H] = -0.125 dex are overplotted as solid black lines for various values of AV, ranging from 2.17 to 2.9 (from left to right), assuming a reddening of E(B-V) = 0.7 mag and a distance modulus of (m-M)V = 11.86 mag. The second and third panels in the upper panel show the FUV-optical and NUV-optical CMDs created using the F154W and N279N passbands, respectively, for the confirmed members of the cluster. Median error bars are displayed in grey on the left side of FUV and NUV-optical CMDs. Bottom panel: CCDs in the FUV-F154W vs F154W, FUV-F169M vs F169M, and NUV-N279N vs N279N planes, respectively. The meaning of the symbols shown is provided in the top-left panel of the first column.
In the text
thumbnail Fig. B.1.

Best SED fits, including single and binary model fits, of the remaining Be stars. The SEDs illustrated in the uppermost panel consist of three stars showing no UV excess. They are well fitted with a single model SED. The symbols and models are the same as in Fig. 1.
In the text
thumbnail Fig. B.1.

Continued.
In the text

Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.

Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.

Initial download of the metrics may take a while.